Ion Exchange Materials || Ion Exchangers, their Structure and Major Properties

CHAPTER 2

Ion Exchangers, their Structureand Major Properties

This chapter begins with very simple ideas, such as ion exchange reactions, their speci-ficity, and the difference between ion exchange reactions generally and the conventionalconcept of ion exchange. The major information in this chapter is a description of pri-mary properties and structure of ion exchange materials. As a result, it provides an ideaabout discussed materials and prepares a reader to follow more advanced chapters.

2.1 Ion Exchange Reactions in Liquid Phase

This section is intended to refresh the memory of a reader who has already forgotten thebasics of chemical interactions. A more advanced reader can simply skip this informationand advance to next sections.

Let us consider homogeneous systems to begin with. Simplest examples of ion exchangereactions can be found in liquid phase. Let us take two solutions: one contains Zn(CN)2,another one contains Hg(NO3)2. Hg(NO3)2 can be considered as completely ionised inaqueous media (see Table 2.1). Practically all Zn2+ ions are initially associated withCN− ions. Mixing these two solutions initiates an ion exchange reaction

Zn(CN)2 + Hg2+ = Hg(CN)2 + Zn2+ (2.1)

This reaction is illustrated by Applet 1.7 Due to much higher stability of Hg(CN)2,the mercury ions replace zinc in the complexes. To explain this, reaction (2.1) can beconsidered as a two-step process:

Zn(CN)2 = Zn2+ + 2CN− log K = logaZn2+a2

CN−aZn(CN)2

= − log βZn(CN)2 = −11

(2.2)

7http://ionexchange.books.kth.se/applet01.html

9

10 Ion Exchange Materials: Properties and Applications

Table 2.1. Constants of 1:2 complexformation (log β ) in aqueous solutions.

CN− NO−3 NO−

3

Zn2+ 11 −0.2 0.85

Hg2+ 34 0.08

Mn2+ 0.80

Hg2+ + 2CN− = Hg(CN)2 log K = log βHg(CN)2 = logaHg(CN)2

aHg2+a2CN−

= 34 (2.3)

where ai is activity of species i; K is thermodynamic constant of the chemical equi-librium; β is the complex formation constant, i.e. constant of the chemical equilibriumresulting in the formation of corresponding complex.8 The constant of ion exchangereaction (2.1) can be calculated as

log KHgZn = log βHg(CN)2 − log βZn(CN)2 = 23 (2.4)

So a high reaction constant secures practically complete replacement of Zn2+ with Hg2+in the complexes.

Interaction involving acetic acid can be taken as another example

(CH3COO)2Mn + Zn2+ = (CH3COO)2Zn + Mn2+ (2.5)

The difference in complex formation constants (see Table 2.1) is not high. Moreover,both complexes are weak. Hence, both zinc and manganese complexes are present in theequilibrium solution as well as free ions of these metals and acetate ions. Behaviour ofions in such systems is illustrated by Applet 2.8a Actual concentrations of all compoundsof the solution can be found solving the following system of equations:

β(CH3COO)2Mn = [(CH3COO)2Mn]γ2

CH3COO−[CH3COO−]2γMn2+[Mn2+] (2.6)

β(CH3COO)2Zn = [(CH3COO)2Zn]γ2

CH3COO−[CH3COO−]2γZn2+[Zn2+] (2.7)

CMn = [Mn2+] + [(CH3COO)2Mn] (2.8)

8Each equilibrium chemical reaction is characterised by corresponding thermodynamic constant expressed interms of activities. However, practical calculations are often performed using concentrations instead. Thisresults in apparent characteristics of chemical equilibria that are truly constant only if activity coefficients ofall involved species are equal to unity.8ahttp://ionexchange.books.kth.se/applet02.html

Ion Exchangers, their Structure and Major Properties 11

CZn = [Zn2+] + [(CH3COO)2Zn] (2.9)

CCH3COO− = 2 · [(CH3COO)2Zn] + 2 · [(CH3COO)2Mn] + [CH3COO−] (2.10)

2 · [Mn2+] + 2 · [Zn2+] = [CH3COO−] (2.11)

where C denotes the total (analytical) concentration of the corresponding substances andγ denotes activity coefficients of corresponding ions. Equations (2.6) and (2.7) reflectstability of two complexes; (2.8)–(2.10) are mass-balance equations; (2.11) represents theelectroneutrality rule. This principle of calculations is widely used to evaluate chemicalequilibria including those involving ion exchange materials.9

A heterophase system can be formed by two liquids that are insoluble in each other.Let us consider a system including aqueous and oil solutions. If the system contains awater-insoluble ligand CH3(CH2)nCOO−, sodium ion, and potassium ion, the followingchemical equilibrium takes place:

CH3(CH2)nCOONa|oil + K+ = CH3(CH2)nCOOK|oil + Na+ (2.12)

The inter-phase distribution can be illustrated by the upper scheme of Fig. 2.1. Thisscheme shows a heterophase ion exchange reaction or ion exchange in the heterogeneoussystem. The apparent equilibrium constant of reaction (2.12) is not far from 1:

˜̃KK

Na =[CH3(CH2)nCOOK

][Na+]

[CH3(CH2)nCOONa

][K+]

≈ 1 ≈[CH3(CH2)nCOONa

][K+]

[CH3(CH2)nCOOK

][Na+]

= ˜̃KNa

K

(2.13)

i.e. none of the ions are preferably accumulated in any of the phases. The double tildeabove ˜̃K indicates that the equilibrium constant is calculated in terms of concentrations (incontrary to the true thermodynamic constant K calculated in terms of activities). Overbarsstand for compounds located in the non-aqueous phase. Deviation of the equilibriumconstant from unity indicates the system where two ions are preferably accumulated indifferent phases. Figure 2.2 illustrates distribution of two ions between phases dependingon the value of ˜̃K. The interactive Applet 410 visualises changes of the distribution ataltering of the constant.

Heterogeneous systems including two liquids, insoluble in each other, are widely usedin solvent extraction separation technique. An ion exchange reaction can be exploited

9If you do not feel confident about interrelations between the constant of the ion exchange reaction andcomposition of the equilibrium solution, experimenting with Applet 3 (http://ionexchange.books.kth.se/applet03.html) can help to develop certain feelings.10http://ionexchange.books.kth.se/applet04.html


K+ Na+

CH3(CH2)nCOONa CH3(CH2)nCOOK

K+

CH3(CH2)nCOONa

water

oil

Co2+

C2H5

C2H5

CH3(CH2)3CHCH2OP

CH3(CH2)3CHCH2O

O

O

Co

O

OH2CH(CH2)3CH3P

O

OH2CH(CH2)3CH3

C2H5

C2H5

water

oil

2 (D2EHPA)H

2H+

Fig. 2.1. Examples of solvent (liquid–liquid) extraction systems. Upper scheme illustrates asodium–potassium distribution between the aqueous phase and oil phase containing a water-insoluble carboxylic acid. Bottom scheme represents a system including Co2+ ion initiallydissolved in the aqueous phase and D2EHPA in the oil phase.

Fig. 2.2. Distribution of two ions between two phases depending on the distribution constantvalue. The dependencies were modelled for reaction AL + B = BL + A at equal amounts of allthree compounds (A, B, and L) in the system. All ligand L was assumed to stay in the oil phasein the form of complexes. Please note that the same plot constructed in a semi-logarithmic scale(C/C0 vs. log ˜̃KB

A) appears as a set of two straight lines.


if the ligand (extractant) forms strong water-insoluble complexes with the desired ion.For example, di-2-ethylhexyl phosphoric acid (D2EHPA)

CH3(CH2)3CHCH2OP

CH3(CH2)3CHCH2O

O

O−H+

C2H5

C2H5

(2.14)

is widely used for extraction of different metal ions. The process of Co2+ extraction canbe discussed as an example. It is illustrated by the bottom scheme of Fig. 2.1. Constantof the chemical reaction is high

K̃CoH = [(D2EHPA)2Co][H+]2

[(D2EHPA)H]2[Co2+] � 1 (2.15)

As a result, all cobalt present is transferred into the oil phase (of course, if there is enoughD2EHPA). This system can be used for extraction of cobalt from water solutions.

May be the main difficulty of the solvent extraction technique is the recovery of theextracted ion from the oil phase. This process, called stripping, follows the extractionstep. When designing a corresponding separation system, one tends to find the most selec-tive ligand to achieve the highest possible separation efficiency, i.e. the chemical reactioncharacterised by the highest equilibrium constant is desirable. However, the increasing ofthe complex stability results in increasing difficulties in the stripping step. The search forsuitable stripping reaction and operation conditions is not always successful. As a result,the solvent extraction systems are often designed compromising the separation efficiencyin benefit of the stripping procedure. Speaking in advance of the following chapters, thesame problem is encountered when a separation is performed using ion exchange materi-als. However, the problem is easily overridden with the use of a column technology thathas a very limited and much less efficient application for solvent extraction processes.

2.2 Ion Exchangers

All reactions and solvent extraction systems, which were described in the previous sec-tion, are based on ion exchange interactions. However, the term ion exchange is mostcommonly applied to interactions including ion exchange materials. The most conven-tional class of such materials is functional polymers: ion exchange and chelating resins.The following text briefly introduces the concept of ion exchange materials with the useof simplest terms and ideas.

A heterogeneous interaction including carboxylic acid CH3(CH2)nCOOH was consid-ered above. The restriction for the hydrocarbon part CH3(CH2)n– was to be longenough, i.e. to make the acid and its complexes water-insoluble but soluble in an organic


non-polar solvent. Acids having large hydrocarbon parts (for example, aliphatic acidsH3C–(CH2)n–COOH: stearic n = 16, arachidic n = 18, or behenic n = 20 acid) aresolid at ambient conditions and can be considered as polymers bearing one carboxylicfunctional group per molecule. Such polymer acts as a second phase when immersed inan aqueous solution. The functional groups remain chemically active and can exchangeions with surrounding media. Of course, only groups, located near the inter-phase bound-ary (surface of the solids) are reachable for ions dissolved in the aqueous phase because,in contrary to the solvent extraction, the groups are immobile inside the phase. The inter-action can be described as an ion exchange involving only functional groups located atthe surface:

RA + B ⇔ RB + A (2.16)11

where A and B are ions; the double-bar indicates the surface location. Reaction (2.16)describes an exchange of two single-charged ions and characterised by the correspondingconstant

˜̃KB

A =[RB

][A]

[RA

][B]

(2.17)

The double-tilde indicates that the reaction constant is calculated in terms of concentra-tions (apparent reaction constant). Charges of ions are omitted in Eqs (2.16) and (2.17)for simplicity. Applet 512 can be used to get a feeling about such interactions and aboutthe effect of the reaction constant.

The dense polymer described above has almost no practical value as an ion exchanger.13

The ratio between the amount of associated ions and the weight of the overall materialis very low. For example, according to stoichiometric calculations,

3H35C17COOH + Fe3+ = (H35C17COO)3Fe + 3H+ (2.18)

one gram of stearic acid can exchange 66 mg of Fe3+ ions.14 However, this theoreticalamount is not practically achievable. Only carboxylic groups located at the surface of thebulk are available for the ion exchange reaction due to the dense nature of this exchanger.

11The ion exchange polymer is often denoted by the symbol R. Symbol R can also denote only the hydrocarbonpart of the organic polyion. In both cases R does not represent the entire polymeric chain but only its structuralunit corresponding to one functional group.12http://ionexchange.books.kth.se/applet05.html13Materials with active functional groups located only on the surface of particles are widely used in analyticalion exchange chromatography (see Section 13.1). However, dense polymers, while being applicable [133], arenot the most advantageous options for such purposes.14The amount of sorbed ions is rarely expressed in milligrams per gram of the material. Milliequivalents pergram (meq/g) or equivalents per kilogram (eq/kg) are conventionally used. This allows obtaining values thatare independent of charge of loaded ions and, hence, to characterise the material itself. Such characterisationwill be discussed in detail in Section 2.4. The calculation result, for example, expressed by Eq. (2.18), is3.5 meq/g, that is similar to that for conventional ion exchange polymers.


Fe3+

Fe3+

Fe3+

COO−COO−COO−

COO−

COO−

COO−

COO−

COO−

COO−

COOHCOOHCOOH

COOH

COOH

COOH

COOH

COOH

COOH

Fig. 2.3. Cross-linked polymer bearing carboxylic groups. The negative charge of groups is com-pensated by hydrogen ions (left) or iron(III) ions (right). Possible associations between iron ionsand carboxylic groups are indicated by dash-lines ovals.

It makes the exchangeable amount almost negligible.15 The solution can be found inuse of a polymer bearing many functional groups in the same chain. Such polymerscan be associated with a large amount of exchangeable ions. Unfortunately, due to thehydrophilic character of the charged groups, such polymeric molecules are water-soluble.To prevent them from dissolving, different polymeric chains have to be cross-linked witheach other forming a three-dimensional polymeric structure. The cross-linking is usuallyperformed with short hydrocarbon bridges. Such a structure is illustrated in Fig. 2.3 whichshows a cation exchanger loaded with two different ions: H+ and Fe3+. Cross-linkedfunctional polymers are not soluble but can swell to a very high degree of the watercontent. Water molecules and ions can migrate within the swollen polymeric network.As a result, most of the groups can participate in ion exchange reactions. Applet 616

illustrates ion exchange interactions in the swollen material and gives an impressionabout the effect of reaction constant.

Few terms should be defined before proceeding to the following sections:

• Ion exchangers are insoluble materials carrying reversibly fixed ions.17 These ionscan be stoichiometrically exchanged for other ions of the same sign.

• Ion exchange polymers, particularly ion exchange resins, are cross-linked polymerscarrying fixed functional groups or sites.

15For example, if we take 0.5 mm spherical beads of stearic acid that is a reasonable size for practical use of anion exchange material and assume that the surface layer thickness is 50 nm, the fraction of molecules availablefor the exchange would be around 0.001, i.e. one gram of stearic acid could exchange only 0.0035 meq of ions(66 µg of Fe3+) that is far below any practical applicability.16http://ionexchange.books.kth.se/applet06.html17There is a group of substances called liquid ion exchangers. Usually they are organic extractants and thusdiscussed within the framework of solvent extraction. As long as these substances are not materials, they arenot subject of this book.


• Functional groups are charged acidic, basic, or chelating groups attached to thepolymer matrix.18 The charge of each group is normally compensated by anexchangeable ion. Inorganic ion exchange materials do not include functionalgroups. Elementary cells of their crystalline structure act instead as the chargedsites whose charges are compensated with exchangeable ions. Functional sites isa wider term than functional groups. It includes the attached groups, chargeableatoms incorporated in polymeric chains, ion exchange units of crystalline structureof inorganic ion exchangers, assemblies of groups in imprinted materials, etc. Mostof the approaches and concepts of ion exchange are equally applicable for differenttypes of functional sites.

• Counterions are exchangeable ions carried by ion exchangers. Counterions canfreely move within the framework but their movement has to be compensated bycorresponding counter-movements of other ions of the same charge to fulfil theelectroneutrality principle. The term counterion has two interpretations:

(i) It can be used exclusively for ions inside the ion exchanger. According to thisdefinition, an ion becomes a counterion only when entering the exchanger.

(ii) It can be used in a broader sense: whether in the exchanger or in the externalsolution, all ionic species with charge sign opposite to that of the functionalgroups can be called counterions.

• Co-ions is a term used for ionic species with the same charge sign as the functionalgroups. The concept of co-ions is not applicable to materials bearing two types ofgroups with opposite charges. These (amphoteric) materials are described below.

• Ionic form is the term defining which counterions are present in the ion exchangerand, hereby, compensates charge of functional sites.

The concept of ionic form is illustrated below. An ion exchanger containing exchangeableNa+ ions, is said to be in sodium form. In the process

R−Na+ + K+ → R−K+ + Na+ (2.19)

the cation exchanger, originally in Na+ form, is “converted” to the K+ form. R inreaction (2.19) and following chemical equations does not represent the entire polymericchain but only its structural unit corresponding to one functional group; the bar indicatesthe exchanger phase. More examples of ionic forms are illustrated in Fig. 2.4. The firststructure represents H+ form of the exchanger. Replacement of hydrogen ions withsodium transfers the material in Na+ form (second scheme). The third scheme illustratesthe case where a double-charged ion compensates charge of two functional groups (Ca2+form). The last is a mixed ionic form that is the usual case in practice of ion exchange.Obviously, the ionic form is not an unchangeable property of the ion exchanger but aterm defining which ion(s) is(are) loaded in the material.

Ion exchangers are conventionally called cation exchangers if they bear negativelycharged functional groups and carry exchangeable cations. Anion exchangers carry anions

18Generally speaking, the term functional groups covers much wider variety of structures. However, acidic,basic, and chelating groups are typical for ion exchange materials.


SO3− +H SO3

− +Na

SO3−Ca2+ −SO3

SO3− +H SO3

− +Na

1 2

3

4

Fig. 2.4. Examples of ionic forms for a polymer bearing sulphonic acid groups. 1 – H+ form,2 – Na+ form, 3 – Ca2+ form, 4– mixed H+, and Na+ form.

due to the positive charge of their fixed groups. The framework of a cation exchangermay be regarded as a macromolecular or crystalline polyanion and that of an anionexchanger as a polycation. If both types of groups are present in the same polymer, it iscalled amphoteric ion exchanger. Some groups are able to form chelate structures withcertain ions and molecules (usually with metal ions). Polymers bearing such groups arecalled chelating polymers or chelating resins. A typical cation exchange was presentedin Eq. (2.19). An example of cation exchange between differently charged cations is

2RNa + CaCl2 ↔ R2Ca + 2NaCl (2.20)

and typical anion exchange is

2RCl + Na2SO4 ↔ R2SO4 + 2NaCl (2.21)

The process (2.20) occurs, for example, in water softening. A solution containing dis-solved calcium chloride (hard water) is treated with a cation exchanger containing sodiumions. The ion exchanger removes Ca2+ ions from the solution and replaces them withNa+. Of course, hard water contains a mixture of Ca2+ and Mg2+ ions and two parallelprocesses identical for both the ions take place.

There is no strict difference between ion exchange resins and chelating resins becausesome polymers can act as chelating or non-chelating substances depending on the


Fig. 2.5. Equivalent character of ion exchange. The following reactions are considered: (a) RNa+K+ = RK + Na+; (b) 2RNa + Ca2+ = R2Ca + 2Na+; (c) R2Ca + 2Na+ = 2RNa + Ca2+.

For illustration purposes the counterions are shown associated with the functional groups that isnot a case for alkali and alkali earth ions in most of the swollen cation exchange materials.

chemical environment. Examples of such interactions will be given in the followingsection.

Ion exchange resembles sorption19 because in both the cases a solid takes up a dissolvedspecies. The characteristic difference between these two phenomena is in stoichiometricnature of ion exchange. Every ion removed from the solution is replaced by an equivalentamount of another ion of the same sign. In sorption, on the other hand, a solute is usuallytaken up non-stoichiometrically without being replaced.20 The competition of two ormore ions for functional sites, and thus precisely stoichiometric character of the process,is illustrated in Fig. 2.5.

To complete the subject, two more definitions of ion exchange and ion exchangers haveto be mentioned [134,135]:

• Ion exchange is the equivalent exchange of ions between two or more ionised specieslocated in different phases, at least one of which is an ion exchanger. The processtakes place without the formation of chemical bonds.

19The term sorption is more suitable than adsorption or absorption. This is because some of the ion exchangeprocesses have adsorption nature, some of them are based on absorption, and, in most of the cases, there isno obvious point of view either the process takes place on the pore surface or in the bulk of the material.20Of course, there are non-exchange sorption processes following strict stoichiometric rules. Furthermore, theion exchange itself can be considered as a kind of sorption interaction. However, the term sorption itself doesnot suggest a necessity of the stoichiometry.


• Ion exchanger is a phase containing osmotically inactive insoluble carrier of theelectrical charge (matrix). Osmotically inactive means that the carrier cannot migratefrom the phase where it is located.21

The first of these definitions can be used to distinguish between classical ion exchangeand chelating interactions because chelating of ions results in the formation of coordina-tion bounds. One could note that reactions involving weak ion exchangers in H+ (cationexchangers) or OH− form (anion exchangers) do not match this definition. For exam-ple, in case of the hydrogen form of a weak cation exchange material (R − COOH),functional groups in the exchanger phase are not ionised. Indeed, reaction

R − COOH + Na+ = R − COO−Na+ + H+ (2.22)

can be considered as a two-step process consisting of ionisation of functional groups inthe internal solution

R − COOH = R − COO−H+ (2.23)

followed by the actual ion exchange

R − COO−H+ + Na+ = R − COO−Na+ + H+ (2.24)

The definition also emphasises the difference between ion exchange interactions andordinary chemical processes, which involve transfer of ions between solution and solidcompounds. For example, transfer of ions between BaSO4 precipitate and SrC12 solution

BaSO4(s) + Sr2+ = SrSO4(s) + Ba2+ (2.25)

is not an ion exchange. This reaction does not match the definition because the newchemical compound SrSO4 is formed and the number of phases is increased (two distinctsolid phases will be present in the system as a result of the reaction) [134].

2.3 Hydrocarbon Structure of Ion Exchange Polymers

Organic ion exchange materials have, as a rule, a three-dimensional polymeric structure.The network is called matrix. Chargeable groups like –COOH, –NH2, etc. are usuallyattached to this “carcass” or directly incorporated in the polymeric chains (–NH– canserve as an example). They can also be attached through short-chain hydrocarbon bridges,called spacers. The overall structure is illustrated in Fig. 2.6.

21Please note that the above mentioned liquid ion exchangers do not fulfil the requirements of this definition,since they do not contain insoluble carrier of the charge [134].


+

−

−

−−

H2O

H2O

H2O

H2O

+

+

+

+

−

H2O

+

++

+

+ +

−

−−

−

−

−

H2O

H2O

H2O

H2O

H2O

1

1

1

1

2

2

3

4

4

5

5

6

6

6

67

7

7

7

(a) (b)

Fig. 2.6. Schematic representation of polymeric ion exchangers. (a) Cross-linked cation exchangematerial; (b) Anion exchange materials with unrecognisable cross-links. 1 – Polymeric chain;2 – cross-link; 3 – physical knot; 4 – negatively charged cation exchange group attached tothe chain; 5 – positively charged anion exchange group incorporated in chains; 6 – counterion;7 – water.

In most of the cases the matrix is constructed from linear polymeric chains cross-linkedwith each other by relatively short links. For example, the above discussed hypotheticalcarboxylic polymer would have the structure:

qpCH

CH

(CH2)r

(CH2)n

(CH2)k

COOH

CH (CH2)m

(2.26)

where left- and right-hand parts represent cross-linking and ion exchange units, respec-tively. Structures of ion exchange materials are irregular in most cases, i.e. theindexes k, m, p, q, and r of structure (2.26) can have any positive integer valuesincluding null. These values can be different for different segments of the polymericchain. The length of cross-linking bridges of the same material is usually the same(n = 1 or 2 or 3 . . . = const) that is defined by the most conventional preparation proce-dure when pre-synthesised and, in many cases, already functionalised polymer is cross-linked with molecules of a monomer, i.e. each bridge is formed from one molecule of


the cross-linking agent. The most typical example of ion exchange resin is sulphonatedpolystyrene cross-linked with divinylbenzene:

CH2CH

n

SO3− H+

CH2CH

CH2CHm

(2.27)

The group –SO3H can exchange hydrogen ion to any other cation, i.e. structure (2.27)belongs to a cation exchanger. Styrene–divinylbenzene matrixes can bear a wide diversityof functional groups. There are different cation exchange, anion exchange, ampho-teric,22 and chelating materials of this type. Styrene–divinylbenzene matrix is the mostconventional because of its chemical and mechanical stability. However, many otherpolymers are also in general use. The most common are copolymers of phenol withformaldehyde

OH

CH2

CH2

SO3H

n

OH

CH2

CH2

OH

CH2

(2.28)

22Many functional polymers contain two or more different types of fixed ionic groups. Such materials are calledbifunctional or polyfunctional. If these groups have opposite charges, the material is amphoteric. Materialsbearing amphoteric groups (groups combining acidic and alkaline units in one regular structure) are alsoamphoteric.


and polyacrylic or polymethacrylic acid with divinylbenzene

C CH2

COOH

R'

n

CH CH2

CH CH2m

(2.29)

where R′ = H in polyacrylic and R′ = CH3 in polymethacrylic polymers. Besides dif-ferent main chains, the diversity of three-dimensional structures is also provided by theuse of different cross-linking agents. While divinylbenzene is the most used agent, manyothers can also be named. Divinyl can be mentioned as a simplest (but not very often)example providing shortest links between chains:

CH CH2

CH CH2(2.30)

A more open gel phase can be achieved with cross-linking agents which are moreflexible than divinylbenzene, for example, trimethylolpropane, trimethylacrylate, andtrimethylolpropane triacrylate [136]. Long flexible cross-linking bridges can providebetter flexibility of the matrix. As a result, ions and water molecules have less difficultyin penetrating inside the material. Thus materials with long cross-links often offer a pos-sibility to exchange large organic molecules which cannot enter inside more conventionalion exchange materials. Sometimes, materials with longer links are referred as macronetion exchange resins [116]. The drawback is an increase in the total weight of the materialwithout increase of its functionality (capacity)23 that is caused by the larger cross-linkingbridges. One has to note that the presence of cross-links does not change properties offunctional groups or, at least, is not supposed to do so. Hence, the replacement of thecross-linking agent usually aims at only altering the structure and mechanical propertiesof the matrix.

Some types of polymers impose special requirements for cross-linking agents. Forexample, preparation of imprinted materials24 requires preservation of the networkconformational structure that is in contrast to the usual non-imprinted ion exchangerswhere cross-linking provides only insolubility and mechanical strength. Ethylene glycoldimethacrylate is remarkable as a cross-linking agent for imprinted polymers. Few morecross-linking agents have been successfully tested for the synthesis of such materials:m- and p-divinylbenzene, triethylene glycol dimethacrylate, butanediol dimethacrylate,methyl methacrylate, etc. [137].

23The concept of capacity will be discussed later in Section 2.6.24Imprinted polymers will be discussed later in Section 3.2.


In most of the cross-linked networks, particularly in styrene–divinylbenzene, cross-linking bridges are very non-uniformly distributed [138–140]. The non-uniformity isdefined by the nature of the radical polymerisation process and can be unfavourablefor practical properties of obtained materials. A uniform distribution can be formedwith the use of other synthetic procedures, for example, by joining the ends of pre-synthesised chains having a narrow length distribution [141,142]. Such materials can becalled isoporous.25

The density of cross-links between polymeric chains is called degree of cross-linking.It influences the structure of the matrix, its elasticity, swelling ability of the material,and mobility of the counterions inside the exchanger, among others. Materials withhigh cross-linking are harder and, in many cases, more stable. However, diffusion insuch materials is slow causing reduction of the rate of all processes. Low-crosslinkedmaterials can be even jelly-like. Their stability could be low in benefit of fast kineticsof interactions. A selection of an appropriate material for some practical application canbe a compromise between desirable stability and reactivity.

Conventionally, the cross-linking degree is expressed as percentage of the cross-linkingreagent introduced in a reaction mixture at the synthesis stage. For example, a well-knownion exchange resin Dowex 50×8 (produced by Dow Chemical for several decades)contains 8% of divinylbenzene molecules. The polymer formally corresponds to thestructure (2.27), however, the cross-linking process can never be perfect. During thesynthesis, a cross-linking molecule could be attached to a polymeric chain with one end,while not being able to acquire an appropriate position for attachment to an another chainwithin the reachable distance. As a result, the structure can be more precisely depicted as

CH2CH

n

SO3− H+

CH2CH

CH2CHm

CH2CH

CH CH2p

(2.31)

The last unit represents the unsuccessfully used cross-linking agent. Nevertheless, theformal cross-linking characteristic – expressed as the percent of cross-linking reagentadded in the synthetic mixture – is very useful because it allows a comparison betweendifferently cross-linked materials of the same type.

Distribution of polymeric chains, cross-links, and functional groups in the bulk of ionexchange resin is not perfect but it is sufficiently even to consider the material as homo-geneous. However, many models consider ion exchangers as two- and even three-phasesystems including the polymer, water inside micropores, and water in macropores.

25Please note that different authors use the term isoporous to emphasise pore uniformity at different scale, i.e.the term could also be applied to pores with size much larger than those defined by the inter-crosslink distances.


Questions of the phase structure of ion exchange materials will be discussed later inChapter 7. In fact, defining the phase of ion exchanger is one of the most difficult tasksof the ion exchange theory. The appropriate way is to consider the material as a singlephase unless the opposite is specifically stated.

While many functional polymers consist of polymeric chains cross-linked with bridgesof regular structure, many others are three-dimensional networks where cross-linkingelements cannot be recognised. For example, the suggested structure of a typicalweak anion exchanger obtained by polycondensation of polyethylenepolyamine withepichlorohydrin/ammonia oligomer can be presented as [143]

CH2

CH

CH2

OH

N

CH2 CH2 NHCH2 CH2 N(CH2 CH2 NH)nCH2 CH2 N

CH2

CH

CH2

NH+Cl−

CH2

CH

CH2

CH2 CH2 NHCH2 CH2(CH2 CH2 NH)nNH CH2 CH2 N

OH

CH2CH

OH

CH2

OH

CH2

CH

CH2

OH

Cl−+HN CH2 CHOH

CH2

....~

....~NH

....~

~....

~....

~....

(2.32)

The anion exchange functionality of this resin is provided with nitrogen atoms. Thescheme (2.32) also illustrates that the functional groups can be incorporated in the poly-meric chains in contrary to binding by side covalent links (as was illustrated by schemes(2.27) and (2.31)).

Besides the cross-linked three-dimensional networks, non-crosslinked polymers mustalso be named. The insolubility and stability are provided to these materials by physicaltangling and knotting of the polymeric chains. Hydrogen bonds and London forces alsocontribute to the insolubility [5]. Such non-crosslinked materials are known but their useis not so wide in comparison with the cross-linked polymers. Of course, the knotting andother interactions (electrostatic, hydrogen, hydrophobic, etc. bonds) between chains alsotake place in cross-linked polymers contributing to their rigidity, pore size, etc. [5,116].Materials with higher number of cross-links also have a higher number of physical knotsand non-covalent interactions that provide even higher density of the net. Chargeablefunctional groups also contribute to the net density, however, this contribution is notinvariable but depends on the state of the groups. When the polymer is dry these groupsare not ionised. As a result, they can attract each other due to dipole–dipole interactions.When ionised, due to swelling of the polymer, functional groups acquire charge andrepulse each other reducing the density of the net. The same effects of the dipole–dipoleattraction and of the ionised species repulsion can be observed in the case of groups thathave two states when surrounded by a solvent: ionised and non-ionised. Most of theother non-covalent interactions between chains also depend on conditions. For example,


an increase in temperature reduces the number of non-covalent links and the net densityapproaches the theoretical structure based on the number of the cross-links successfullyformed during the synthesis.

A special class of materials is hypercrosslinked polymers [144].26 They are preparedby extensive cross-linking of linear chains in a strongly solvating media. Generally,the basic principle of obtaining hypercrosslinked polymers consists of the formationof a rigid network in highly solvated state. Owing to the high rigidity and reduceddegree of chain entanglement, such “expanded” networks are characterised by a loosechain packing, i.e. high free volume, and by the unique ability to swell in both polarand non-polar media. These materials are relatively new, however, their advantages inboth analytical chromatography [145] and separation technologies [146] can be foreseen.These advantages can be specifically high for extraction, separation, and purification oforganic substances.

Ion exchangers called snake in cage materials must also be mentioned. They consistof three-dimensional polymeric networks and physically trapped large (often linear)molecules of organic electrolytes [147,148]. The principle illustrating the name of suchmaterials is shown in Fig. 2.7. Of course, two “parts” of such materials are rather inter-connected by tangling and knotting of chains than by trapping in cavities. Snake in cageion exchangers are usually modified ion exchange resins consisting of a cross-linkedpolymeric network with fixed charges and trapped linear polyelectrolyte of the opposite

C00

H

Fig. 2.7. Illustration of snake in cage phenomenon. The electrolyte is trapped inside the networkbut the ionisable group (carboxylic group in this illustration) is available for interactions withspecies dissolved in the internal solution of the material.

26Hypercrosslinked polymers are also referred as macronets.


charge [149].27 The preparation can be done with a few different approaches. One of theapproaches begins with an appropriate ion exchange resin loaded with inorganic ions.The ionic form is selected to get maximum swelling of the material.28 Loading of thematerial with organic ions and the removal of inorganic ions causes a shrinking of theswollen polymeric structure. The shrinking process usually takes longer time than the ionexchange reaction and, thus, the time is sufficient to trap the organic species inside thematerial. The reversed reaction cannot be performed because the shrunk ion exchangerdoes not allow free diffusion of trapped species inside large organic species. As a result,only small inorganic ions can penetrate the exchanger phase. Materials of this type arealso prepared by polymerisation of a “snake” monomer inside the structure of an ionexchange resin. An example is polymerisation of acrylic acid inside the anion exchangeresin Dowex-1 [150]. Two oppositely charged functional sites of the same material are inintimate contact and tend to neutralise each other, however, the sites still have an attrac-tion to cations and anions and tend to associate with them to some extent. The result isthat the material sorbs both anions and cations from surrounding solutions. Equivalentamounts of anions and cations can be sorbed simultaneously. The ions, except for H+,are held so weakly that they can be eluted with water [150]. Such materials can alsoexchange only anions or cations depending on the composition of the liquid phase [151].More about materials bearing two opposite chargeable types of groups are discussed inthe next section.

2.4 Functional Groups

The main difference between polymeric ion exchangers and non-functional polymers isthe presence of functional groups or functional sites in the structure, i.e. the presence offunctional units is the main feature of the materials discussed. There is a wide diversity offunctional groups that have been attached to polymeric networks to obtain ion exchangeproperties. Different examples are given in Table 2.2. However, this diversity is reducedto only a few types of groups when marketed materials are listed. The reason could ratherhave a commercial than a chemical origin.

Chemical properties of functional groups define the type of a particular ion exchangematerial. Cation and anion exchangers are the two types bearing respectively the neg-atively and positively charged groups and, hence, being able to exchange cations oranions. Due to different dissociation properties of groups, strong and weak exchangersare recognised similar to that of strong and weak electrolytes.29 Table 2.2 also pro-vides an idea about dissociation constants of some groups if associated with H+ (cation

27Polymeric materials containing two oppositely charged types of groups with first type covalently boundedand second type physically trapped are also called ion retardation resins.28Questions of swelling will be discussed in detail in Chapter 7. The important property at this point is thatthe swelling degree of ion exchange resins in the form of organic ions is usually much less than those in theform of inorganic ions.29For example, acids that are completely ionised in aqueous solutions are considered to be strong. Similarly,the group –SO3H being an analogue to the sulphuric acid, is completely ionised in the internal solution of thematerial. Ion exchangers bearing these groups are strong cation exchangers. Weak ion exchangers are oftenconsidered as materials having working ranges of pH, i.e. being able to exchange ions only if pH allowsionisation of their functional groups.


Table 2.2. Examples of functional groups and corresponding types of the ion exchangematerials.a Dissociation constants, pKa, are given in parenthesesb [71,112,116,151–155].

Cation exchange materials; negatively charged groups

–SO−3 (≤1) –PO2−

3 (2.0–4.5; 6.5–9.5)

–COOH (3.5–8) −HPO−2

–C6H4OH (≈10) –SH (>10)

–AsO−3 –SeO−

3

Anion exchange materials; positively charged groups

–N+(CH3)c3 (≤1) −N+(CH3)2C2H4OHd (1–2)

–N+(CH3)2C2H5 −P+(C4H9)3

≡ N (6–9) N+

(2–3)

N (9–10) N C2H5

+

N C2H4 CO CH3

+N C2H4 CO NH2

+

Amphoteric (bipolar) materials; both type of groups: negatively and positively charged

–SO−+3 H and −N(CH3)2 –COOH and −N+(CH3)3OH−

Amphoteric materials; bipolar groups

N N

NN

CCH2 SCOOH

+N N

NN

CCH2 S SO3−+

Chelating materials; fixed groups which are able to form chelate rings with metal ions

NCH2COOH

CH2COOHCH2

a As mentioned earlier, the distinction between the different types of ion exchangers can be quite arbitraryin many cases. This is specifically true when one speaks about the difference between amphoteric andchelating materials.b Dissociation constants are presented here for some groups in order to provide the reader with an idea aboutthe difference in reactivity of these structures. The constants correspond to associates with H+ or OH−ions. These values can be compared only within semi-quantitative considerations because they have beendetermined by different methods under different conditions. Moreover, theoretical discussions on the natureof the groups dissociation are not over yet and there is no conventional approach to this characteristic.c Materials bearing −N+(CH3)3 groups are conventionally called Type I anion exchangers.d Materials bearing −N+(CH3)2C2H4OH groups are conventionally called Type II anion exchangers.


exchangers) or OH− (anion exchangers). These constants cannot be considered as ref-erence quantities because their values are dramatically dependent on the determinationmethod and on the selected physico-chemical model. However, the table gives a goodidea for semi-quantitative comparison of strength exhibited by different groups.

The most common cation exchangers are strongly acidic resins with sulphonic acidgroups (−SO−

3 ) and weakly acid resins with carboxylic acid groups (–COO−). Differention exchangers bearing the same groups can exhibit significantly different propertiesbecause the strengths of the groups depend on the nature of the supporting hydrocarbonstructure. Examples of this dependency are presented in Table 2.3. However, to obtaina wide diversity of materials matching different requirements, other types of cationexchangers have also been developed. These materials carry phosphonic, phosphinic,arsonic, selenonic acid groups as shown in Table 2.2. Most of the functional groups ofanion exchangers contain nitrogen as a proton-accepting atom. Similar to low molecular

weight amines, weak-base (−NH+3 , > NH+

2 ) and strong-base (for example, −|

N+| − or

−N+(CH3)3) anion exchangers are recognised. Besides the nitrogen-containing units,

strong-base quaternary phosphonium (−|

P+| −) and tertiary sulphonium (−

|S+ −) anion

exchange groups are known.

Some scientists do not consider chelating polymers as ion exchange resins because ofdifferent phenomena defining their affinity to ions. These materials also do not fit someclassic models of ion exchange systems that will be discussed in detail in Chapter 7.However, depending on the chemical system, many chelating groups can act in preciselythe same way as ion exchange groups without exhibiting any chelating property. Forexample, an aminophosphonate chelating resin can act as a conventional non-specificion exchanger or as a chelating polymer [156]. These interactions are illustrated inpanel (a) of Fig. 2.8. The first (upper) branch of the scheme illustrates ion exchangebetween sodium and potassium ions

R2−(Na+)2 + 2K+ = R2−(K+)2 + 2Na+ (2.33)

that is the most proverbial example of a pure ion exchange. Precisely the same reactioncan take place in any cation exchanger (only the meaning of R in Eq. (2.33) would bechanged). No direct association between the functional group and corresponding ions of

Table 2.3. pK of functional groups attached to different hydrocarbons.

Hydrocarbonradical

Functional group

–SO3H –COOH –PO3H2 –NH2 –NHCH3

H3C– 1.92 4.8 8.0 2.2 10.6 10.8H5C2− 1.68 4.9 – – 10.6 –H3C–(CH2)2− 1.53 4.8 – – 10.6 –Ar– −0.04 4.2 7.5 1.9 4.6 4.8p–H3C–Ar– 0.03 4.4 7.7 2.0 5.1 –


(a)

(b)

R

CH2

NH

CH2 P

OOHOH

R

CH2

NH

CH2 P

OZn2+

R

CH2

NH

CH2 P

OO−Na+ Na+

R

CH2

NH

CH2 P

OO−O−

K+ K+

−2Na+

−2Na+

+Zn2+

+2H+

+2K+−2Na+

−2Na+

−2Na++2H+

+2K+−2Na+

CH2 NCH2COOH

CH2COOHR

Cu2+

CH2 NCH2COO−Na+

CH2COO−Na+R

CH2 NCH2COO−K+

CH2COO−K+R

CH2 NCH2COO−

CH2COO−R

+Cu2+

O−

O− O−

Fig. 2.8. Different interactions involving aminophosphonate (a) and iminodiacetic (b) groups: ionexchange (upper branch in each panel), protonation (middle branches), and chelating (bottombranches).

opposite charge happens. The interaction perfectly fits the classic model of ion exchangematerial, where counterions are distributed in the internal solution of the ion exchangerand no direct interaction between them and functional sites is supposed. The secondbranch illustrates protonation of the same aminophosphonate group

R2−(Na+)2 + 2H+ = RH2 + 2Na+ (2.34)

Two hydrogen ions replace two sodium ions. The reaction is typical for weakly acidiccation exchange resins. It is also considered as an ion exchange interaction. The third


example is the replacement of sodium ions with zinc and the formation of a chelatestructure. Such reaction is typical for chelating polymers. It is perfectly stoichiometricand can be described by the same type of ion exchange reaction:

RNa2 + Zn2+ = RZn2 + 2Na+ (2.35)

The similarity between reactions (2.33)–(2.35) cannot be overlooked. They also canbe described using the same thermodynamic approaches. Another example is the imin-odiacetic group which is well-known for strong complexes with transition metals, forexample, with Cu2+. However, it acts as a conventional ion exchanger in systemsthat include only hydrogen and alkali metal ions. The corresponding interactions areillustrated in panel (b) of Fig. 2.8.

The −PO2−3 is conventionally considered as a cation exchange group. However, it can

serve as an inverted example. Two of such groups can form chelate complexes withfour-charged metal cations [116]

Me4+P

O−

O−

O

CH

P

O−

O−

O

CHCH2

CH2

(2.36)

Besides functional structures presented in Table 2.2, many other sophisticated units canbe useful for some specific applications. For example, an advantage of chelating groupshaving a long spacer between chelating atoms and the polymeric matrix was demonstratedby comparison of Zn2+ and Cu2+ complex forming ability of three groups [157]:

R − CH2 − NH − C||S

− NH2 (2.37)

R − CH2 − NH − (CH2)2 − NH − C||S

− NH2 (2.38)

andR − CH2 − NH − (CH2)3 − NH − C||

S

− NH2 (2.39)

The groups (2.38) and (2.39) were found to chelate with copper stronger than with zincdue to their higher local mobility and, hence, accessibility for chelating interactions.There are reports that the introduction of spacers between the benzene rings of polymericnetwork and quaternary amines also improves chemical [152,153,158], γ-ray [158], andthermal [158,159] stability of strong anion exchangers. Enhanced reaction rates wereobserved for spacer-modified anion exchangers [159].


2.5 Inorganic Ion Exchangers

Besides organic polymers bearing functional groups, many inorganic materials exhibitability for the stoichiometric exchange of ion. Functional polymers are the first choicefor ion exchange materials because of their greater mechanical and chemical stabilityin different media as well as reproducibility in behaviour. However, the attention isconstantly brought to inorganic materials due to their much higher stability at elevatedtemperatures and in the presence of a strong radiation. These special features of inor-ganic ion exchangers are significant for their use in radioisotope separations and solvingproblems of nuclear waste [160]. Inorganic ion exchangers are also highly advantageousin a number of catalytic applications [161].

While the chemical composition of inorganic ion exchange materials can be highlydiverse, they are typically single or mixed metal oxides, hydroxides and insolubleacid salts of polyvalent metals, heteropolyacid salts, and insoluble metal ferro-cyanides [38,160]. The wide diversity of such materials includes natural minerals (“as is”and modified) and synthetic substances. They can be amorphous but, more often, areinorganic crystalline polymers with microporous framework structure. A short list ofinorganic ion exchange materials would include zeolites, titanates [162], silicotitanates[163–165], hydrous oxides [166–168], phosphates of zirconium and other transitionmetals [166,167], tin(IV) compounds [166,168], manganese compounds [166,168], hex-acyanoferrate compounds [165,169–171], hydroxides, and oxides, among others. Ionexchange properties of all these materials are defined by cavity-containing or layeredstructures with exchangeable cations located on internal surfaces of these voids.

Zeolites are the most conventional inorganic ion exchangers, occurring naturally but alsosynthesised in a wide diversity of variations. They are hydrated aluminosilicates withcrystalline porous structure. Examples of their crystalline morphology are presented inFig. 2.9. Generally, the structure can be regarded as an inorganic polymer built fromtetrahedral MeO4 units, where Me is an Si4+ or Al3+ ion [173]. Each oxygen atom ispositioned in a corner of the unit and shared between two Me atoms, thus forming aninfinitely extended three-dimensional network. The chemical composition of zeolites isusually represented by the following empirical formula

A2/ZO · Al2O3 · ySiO2 · WH2O (2.40)

where A is the counterion with valence Z; y accounts for the SiO2/Al2O3 ratio; and W isthe water content in the hydrated form of the zeolite. The value of y is usually between 2and 10 but can be up to 100 in “high-silica” zeolites [173]. Please note that the amount ofcounterions in the structure stoichiometrically corresponds to the number of aluminiumatoms. Formula (2.40) reflects stoichiometry of the material that can be expressed as

AZ+ : Al3+ : Si4+ = 2/Z : 2 : y (2.41)

The representation (2.40) could be somehow confusing because it shows an assem-blage of oxides and, as a result, gives an impression of a chemically inert mixture. Toexplain the principle, let us consider the material as a crystalline structure with generalformula (MeO4)m. Left panel of Fig. 2.10 shows such structure consisting solely of SiO2


Fig. 2.9. Examples of zeolite crystals: (a) single crystals of zeolite A; (b) single crystals of anal-cime; (c) single crystals of natrolite; (d) zeolite L; (e) typical needle aggregates of zeolite mordenite;(f) typical needle aggregates of Nu 10. Reprinted from J. C. Jansen, in: Introduction to ZeoliteScience and Practice [172] 1991, Elsevier ©.


Si4+

O2−

O2−

Si4+

Si4+

O2−

O2−

O2−

O2−

Si4+

Si4+

O2−

O2−

O2−

O2−

Si4+

Si4+

O2−

O2−

O2−

O2−

Si4+

O2−

O2−

O2−

O2−

Si4+

O2−

O2−

Si4+

Al3+

O2−

O2−

O2−

O2−

Si4+

Si4+

Al3+

O2−

O2−

O2−

O2−

Si4+

Si4+

O2−

O2−

O2−

O2−

Al3+

O2−

O2−

O2−

O2−

Si4+

Fig. 2.10. Non-charged structure of silica (left panel) and charged structure of a zeolite (rightpanel). Counterions that are necessarily present in the Al-containing structure are not shown forsimplicity.

units (structure of silica). All positive charges provided by Si4+ ions are compensatedby negative charges of O2−. According to expression (2.41), such a structure does notinclude counterions. The specific ion exchange nature is provided by a partial replace-ment of Si4+ with Al3+ (right panel of Fig. 2.10). Such replacement results in an excessof negative charge that is necessarily compensated by the presence of counterions thatare not included in the well-structured oxide units. As a result, counterions are heldin crystalline pores and do not occupy precisely defined crystalline positions. Thus therepresentation more convenient in the framework of ion exchange is

AZ+2/Z[(SiO2)y(AlO2)

−] · WH2O (2.42)

There are several tens of different zeolite structures: 40 to 50 different natural mineralsand numerous synthetic forms. Figure 2.11 provides an idea about structures of thesematerials that are frameworks with voids interconnected by channels. The most commonchannel cross-size is between 4 and 7 Å but can be up to 20 Å [173]. The pore sizeis conventionally discussed as the outcome of the number of oxygen atoms formingthe structure around each channel. As illustrated in Fig. 2.12, this number is differentfor different zeolites while being unchanged within the same type of structure. Preciseregularity of pore dimensions makes such materials essentially different from the abovedescribed organic ion exchangers that are irregular in most cases.

As is obvious from Fig. 2.10, the crystal structure of zeolites consists solely of silicon,aluminium, and oxygen atoms. The counterions and water molecules are located insidethe channels. As long as counterions are not incorporated in the crystalline structure butstay inside the material to fulfil the electroneutrality law, they can be exchanged forother ions of the same charge. The exchange is stoichiometric and reversible that makeszeolites fully capable ion exchange materials.

Due to the regular zeolite structure, all tetrahedra metal-oxygen units are exposed tothe surface of pores and thus are accessible. Moreover, the surface is formed by atomswith unchanged coordination, that is in contrast to the surfaces of bulk materials that areformed by the breakage of bonds [176]. The only exception is the outer surface of the


Fig. 2.11. Three-dimensional crystal structure of erionite. Reprinted from H. Kalies, F. Roessner,H. G. Karge, and K.-H. Steinberg, in: Zeolite Chemistry and Catalysis [174] 1991, Elsevier ©.

Fig. 2.12. Examples of 3-connected two-dimensional nets observed in zeolites. Reprinted fromH. van Koningsveld, in: Introduction to Zeolite Science and Practice [175] 1991, Elsevier ©.


zeolite crystals where the structure is terminated [176]. Besides excellent ion exchangeproperties, such porosity provides easiness for modifications such as replacement ofSi and Al in the framework. The modifications lead to altering of different propertiesincluding preference to incorporation of different counterions [176]30. This, togetherwith a possibility to obtain a high diversity of channelled shapes, allows to designadvanced ion exchangers and ion exchange sorbents.

Ion exchange reactions with zeolites and other inorganic exchangers can be consideredwith the same approaches as reactions involving organic materials. However, a few spe-cific advantages and disadvantages can be recognised. Natural zeolites do not exhibit ahigh preference to one kind of counterions, however, some synthetic and modified mate-rials as well as some other inorganic exchangers can be highly selective [160,177–180]and, hence, can allow to design highly specific ion exchange processes. For example,hexacyanoferrates and some salts of heteropolyacids are known to possess high affin-ity for caesium. Hydrous metal oxides, titanosilicates, and titanates have shown a highpreference for strontium ions [180].

Due to the possibility of altering the Si/Al ratio in zeolites, their properties can be adjustedfor specific needs. For example, an increase in this ratio provides a higher thermalstability, resistance to acidic environment, and hydrophobicity, whereas the ion exchangecapacity decreases.

The uniform pore dimensions allow certain ions to enter the crystals while rejecting theothers. This very precise discrimination of chemical species on the basis of molecularsize is the reason that zeolites are conventionally regarded as molecular sieves. Thecorresponding effect is illustrated in Fig. 2.13. This phenomenon combined with chemicalpreference of the material allows designing of highly specific processes, for example,selective catalytic conversion of small organic molecules while larger homologs remainunchanged.

As a rule, the inorganic exchangers are cation exchange materials. However, someof them, for example, hydrous tantalum phosphates possess amphoteric property. Thisbehaviour can be depicted in Fig. 2.14. The difference can be easily recognised betweenamphoteric behaviour of tantalum phosphates and amphoteric organic ion exchang-ers (described in Section 3.5). Such organic materials possess two different types offunctional groups attached to the same polymeric network.

One of the main problems in use of inorganic exchangers for separation purposes isa narrow pH range of operation because of the dissolution of aluminium and siliconfrom the framework at harsh pH conditions. Another problem is that a quantitativereplacement of counterions by means of auxiliary reagents is difficult and, in some cases,impossible [178]. Relatively pure reproducibility of interactions involving inorganic ionexchangers, and the difficulty for quantitative elution of sorbed ions can be, among othercauses, the results of pH limitations. As known from the procedures involving organicion exchangers, most conventional way to regenerate them for a repeated use is an acidic

30Questions of preference to one or several kinds of ions, so called selectivity, will be discussed later inChapter 8.


Fig. 2.13. Illustration of molecular sieve effect. Straight chain molecule of normal octane (top)passes through eight-ring aperture of 5A (CaA) zeolite; branched molecule of iso-octane (bottom)cannot. Reprinted from E. M. Flainigen, in: Introduction to Zeolite Science and Practice [173]1991, Elsevier ©.

OH

+ H2O

+ Ct1+OH−

O−Ct1+

OH2+An1

−OH2

+An2−

+ H+An1

− Ct1+

O−Ct2+

+ Ct2+

+ An2+

− An1+

Fig. 2.14. Ion exchange interactions involving hydrous tantalum phosphates [168]. Two branchescorrespond to the behaviour in different media: anion exchange in acidic medium (upper branch)and cation exchange in alkaline medium (bottom branch). Cti and Ani indicate cation and anionrespectively. Exchange of single-charged ions is shown for simplicity.


treatment that is inapplicable to most of the inorganic exchangers. For example, an acidicform of a zeolite cannot be obtained by an ion exchange reaction. The material is loadedwith ammonium instead

R−Ct+ + NH+4 = R−NH+

4 + Ct+ (2.43)

The following heat-treatment converts the material in the acidic form

R−NH+4 = R−H+ + NH3(gas) (2.44)

R− in Eqs (2.43) and (2.44) represents charged crystalline structure of the inorganiccation exchanger.

Another drawback is a relatively low rate of ion diffusion inside the crystalline inor-ganic materials (much lower than in organic ion exchangers). This affects the overallperformance of the material and may offset the economic advantages [38].

2.6 Ion Exchange Capacity

Due to the equivalent character of ion exchange and the defined number of functionalgroups in the ion exchanger, an ion exchange reaction can be considered as a competitionbetween counterions for functional groups of the material. This competition can bewritten as

[A] + [B] + · · · = Q (2.45)

where concentration of ions is expressed in the equivalent concentration scale as wellas the total amount of ions inside the exchanger, Q, so called capacity. Equation (2.45)defines the primary difference between the exchanger phase and conventional solutionswhere the total concentration is limited by only the solubility of the constitutes.

Ion exchange capacity is a major characteristic of ion exchange materials. From a prac-tical point of view, an ion exchanger can be considered as a “reservoir” containingexchangeable counterions. The counterion content in a given amount of material isdefined essentially by the amount of fixed charges which must be compensated by thecounterions, and thus is essentially constant. According to this fact, ion exchangers arequantitatively characterised by their capacity which is defined as the number of counterionequivalents31 in a specified amount of the material [5].

The above definition gives a good filling about the chemical nature of the capacity.However, it does not reflect either physico-chemical or practical property of any ion

31Equivalents are units “legally” condemned in favour of moles because, in most chemical systems, thenumber of equivalents is simply defined as the number of moles multiplied by the valence. On the contrary,the amount of functional polymers cannot be expressed in moles. An equivalent represents the amount of thematerial providing Avogadro’s number of reacting units, i.e. the weight of one equivalent equal to the weightof the polymer fragment bearing one functional group multiplied by 6.022·1023. Similarly, the Normality (N)concentration scale is still used for functional polymers instead of Molarity (M).


exchange material. Scientists and engineers use many different definitions of capacitydepending on practical need and even individual preferences. As a result, each materialcan be characterised by several values of capacity. Even more, the conditional capacity(capacity exhibited under certain particular conditions) is often used to describe andcompare ion exchangers. Taking the infinite number of possible operating conditions,the number of capacity values for one material can be raised up to infinity.

The difficulty in defining capacity has two causes. First, the availability of functionalgroups for exchange reaction never achieves 100%. Different fraction of the groupsparticipates in each process depending on the macrostructure of the material, the swellingdegree, and the size of exchanged ions, among others. Surprisingly, the second problemis the difficulty in defining the amount of the material. This is related to the mainfundamental problem of ion exchange, that is uncertainty of the boundary between thephase of ion exchanger and the phase of surrounding media. The difficulty is encounteredin the selection of which part of the water should be counted as an essential part of thematerial.32 Moreover, the swelling varies depending on the conditions (see Chapter 7) thatis reflected by the difference in values of conditional capacities. Capacity data suppliedby manufacturers occasionally refer to “air-dry” material,33 i.e. to the material containingan indistinct amount of water. Such data characterise value of the marketed product ratherthan the properties of the ion exchange material. Another question related to the amountof the material is how the presence of counterions should be taken into account. In mostof the cases, the H+ form of cation exchangers and Cl− form of anion exchangers areselected as standard, i.e. the capacity is calculated per weight of the material includingthe weight of sorbed H+ and/or Cl− ions. The Cl− form has been selected despite thelogical use of OH− form because of certain difficulties in handling the hydroxide forms(see Appendix I). However, in many particular cases, capacity is calculated for other ionicforms. The outcome of these uncertainties is the fact that the capacity data stated withoutany description of how they were obtained should be accepted with reservations [5].

The most precise characteristic of capacity is, most probably, least useful for estima-tion of the material practical performance. The number of functional groups per weightunit of the dry ion exchanger34 is defined as theoretical capacity or as content of func-tional groups. This characteristic can be obtained directly by analysing the material.For example, each atom of sulphur corresponds to one functional group in a sulphonatedpolymer (for example, see Scheme (2.31)), hence the number of functional groups canbe derived from the sulphur content

QTheor = mS

32 · mr · (1 − W)(2.46)

32The problem is addressed in more detail in Chapter 7.33Air drying usually means keeping the exchanger in an open-air container for a long time. So far as thewater-content of the exchanger is sensitive to the air humidity, air-dry weight is a very imprecise characteristic.34Dry weight of an ion exchanger can also be defined in different ways. The uncertainty comes from watermolecules hydrating functional groups. For example, each sulphonic group of styrene–divinylbenzene ionexchangers is strongly associated with one water molecule. The fixation of this molecule is so strong thateven heating up in air rather initiates destruction of the matrix than evaporation of this water. As a result, thereference to dry weight usually means “weight of dried material”. Nevertheless, the dry weight is a value thatis easy to standardise.


where mS is the amount of sulphur, mr is the amount of the polymer taken for theanalysis, W is the water fraction in the sample, 32 is the atomic weight of sulphur. Thevalue of QTheor can also be estimated from the synthesis conditions. The theoreticalcapacity has no practical significance, however, it indicates the limiting value, i.e. thetheoretical maximum that can be approached but can hardly be achieved. No other ionexchange capacity can exceed the theoretical capacity.35 The theoretical capacity isusually measured in meq/g (milliequivalent per gram) of dry material.

Any practically valuable characteristic of ion exchange capacity can be defined only asa content of the counterion in the ion exchanger, i.e. it is a ratio between the amount ofions and the amount of the substance containing these ions:

Q = mion · Zion

mion exchanger(2.47)

where Zion is the ion charge. As was already explained, there is no one standardised wayto define the amount of the ion, mion, and the amount of the ion exchanger, mion exchanger .Both definitions are strongly related to a problem to be solved. The most commondefinitions for mion are:

• number of functional groups available for any kind of ion exchange or number ofexchangeable counterions;

• number of functional groups available for certain kind of ions or maximum numberof this kind of ions which can be accommodated;

• number of functional groups utilised for sorption of certain ions in certain processor number of this kind of ions that was actually sorbed in the process;

• number of functional groups occupied by certain ion at equilibrium state undercertain conditions or number of this kind of ions sorbed at equilibrium state underthese conditions;

• total number of functional groups (this definition relates to the theoretical capacityand presented here to complete the list).

The most common definitions for mion exchanger are:

• dry weight of the material;

• air-dry weight of the material;

• volume of the swollen material;

35There are chemical phenomena leading to sorption of ions in excess of the theoretical capacity. However,these phenomena cannot be considered as purely ion exchange interactions.


• volume of the exchanger bed36 including the swollen material and the solutioncontained in inter-bead voids;

• weight of the solvent contained in internal micropores of the material;

• volume or weight of the solvent contained in the bed;

• one unit of the ion exchange equipment, for example, an overall ion exchangecolumn.

When reading scientific and technical literature, one can find all possible combinationsof these two (mion and mion exchanger) to define ion exchange capacity. Besides the totalnumber of functional groups per dry weight, defining the theoretical capacity, mostcommonly used combinations are:

• Total number of functional groups per volume of the swollen material. This char-acteristic can be identified as concentration of functional groups in the material.The theoretical capacity and the concentration of functional groups can charac-terise results of the material synthesis. However, they are hardly useful to reflectthe exchange performance because certain part of the groups is not available for ionexchange. These groups are completely surrounded by hydrocarbon chains of thematrix. The counterion allocated near such a group has no way to diffuse out of thephase and a replacing ion has no way to go in.

• Total ion exchange capacity is usually defined as the number of functional groupsavailable for any kind of ion exchange or number of exchangeable counterions percertain amount of the material. In practice, the maximum achievable amount ofsorbed ions is referred when the total ion exchange capacity is reported. The totalion exchange capacity is usually calculated per weight of the dry material.

• Number of functional groups available for a certain kind of ion or maximum numberof this kind of ions which can be sorbed per certain amount of the material is capacitytowards certain ion (for example, capacity towards Ca2+). In some instances it iscalled total capacity towards the ion. This definition is used when the interactionswith ions of particular interest are considered. Capacity towards the ion can differfor different ions due to different ionic size and, as a result, different ability of ionsto penetrate dense regions of the ion exchanger. This characteristic is calculated perweight of the dry material or per volume of swollen material.

• Number of functional groups per unit of internal solution contained in the material.This characteristic is rarely referred as capacity. The more common term is

36Volume of a reactor (for example, an ion exchange column) packet with beads or particles of an ion exchangematerial is referred to as ion exchange bed. The fraction of the volume of the bed that is not filled with the ionexchanger (inter-bead voids) varies between 0.25 and 0.55 depending on the shape and size of the exchangerbeads [112].


concentration of functional groups in the internal solution.37 The number of groupsavailable for ion exchange is more commonly used than the total number of groups.

• Number of available functional groups or number of exchangeable counterions perunit of fully swollen packed bed is referred often as volume capacity or technicalvolume capacity. In most cases, the volume of the settled bed is used, however, thevolume under the operating conditions is another common case.

• Number of functional groups utilised for the sorption of certain ions in certainprocess or number of this kind of ions which was actually sorbed in the process canbe called utilised capacity. The most common denominator for the utilised capacityis the volume of the bed including the swollen beads and inter-bead solution. Thevalue reflects performance of practical separation processes. This value is stronglydependent on construction of the reactor, operating conditions, etc. Sometimes thisvalue is also referred as useful capacity.

• Number of functional groups occupied by certain ion at equilibrium state undercertain conditions or number of this kind of ions sorbed at equilibrium state underthese conditions is called equilibrium capacity. It is calculated per weight of thedry material or per volume of the swollen material. The term equilibrium capacityis strictly linked to the conventional method of characterising, where ion exchangematerials are investigated under standardised equilibrium conditions.

These “capacities” are characterised by differently selected mion exchanger. The selectionis due to the following:

• In scientific research, this amount is calculated per dry weight of the material, or pervolume of the swollen material, or per amount (weight or volume) of the solventcontained inside the beads because the calculation aims to characterise the material.

• Commercial information deals with air-dry weight because the air-dry product isusually marketed.

• Volume of the bed is commonly used in chemical engineering and chemicaltechnology because the overall performance of the bed is important.

One has to understand that, from a strict point of view, the terms utilised capacity andequilibrium capacity are misleading. The correct words are loading and equilibriumloading, i.e. the amount of the matter loaded into the exchanger. However, the wordcapacity is widely used in these instances by scientists and chemical engineers.

The amount of exchangeable ions per overall reactor has to be mentioned separately asion exchange capacity of the reactor or ion exchange capacity of the column. This valuehas nothing to do with performance of the material but it characterises performance ofthe industrial unit. For example, the following columns have the same ion exchangecapacity: one loaded with 1.00 m3 of a cation exchanger with capacity 2.0 equivalents

37Please note, when speaking about concentration of functional groups, the inter-bead void is not included inthe calculation that is in contrast to the technical volume capacity which refers to unit volume of packed bedincluding the interstitial void.


per litre and another loaded with 0.80 m3 of a material with capacity 2.5 equivalents perlitre (1.0 m3 · 2.0 eq/L = 0.80 m3 · 2.5 eq/L).

Another technology concept is breakthrough capacity. If the ion exchange materialis used for a column process (see Chapter 11), the operation is often discontinued atbreakthrough, i.e. when the ion or substance that is targeted to be trapped by the columnappears at the outlet. As a result of such process, the loading of material is not homoge-neous along the longitudinal axis of the column. In case of imperfect performance of thecolumn, a radial inhomogeneity can also occur (see Section 11.4). As a result, the valueof breakthrough capacity of the column is always lower than that of an ion exchangecapacity characterising the material loaded in this column. The breakthrough capacitydepends on the operating conditions and on the properties (including capacity) of thematerial, however, the reversed calculation is impossible, i.e. the breakthrough capacitycannot be used to obtain the material’s characteristics.

Non-exchange sorption of ions and molecules by ion exchange materials often occurstogether or instead of ion exchange reactions. Such phenomena can be caused by manydifferent factors that will be described later. Two general types of non-exchange sorptioncan be recognised: stoichiometric and non-stoichiometric. Examples of stoichiometricprocesses could be sorption of anions on non-protonated weak-base anion exchangegroups from acidic solutions [5,181]:

R − NH2 + H+ + Cl− = R − NH+3

−Cl (2.48)

and sorption of zwitterions on H+ forms of cation exchangers

R−+H + H3N+ − R′ − COO− = R−+NH3 − R′ − COOH (2.49)

where R′ is hydrocarbon of the zwitterion. Ion exchange reactions similar to (2.48) and(2.49) can be easily written:

R − NH+3

−Br + Cl− = R − NH+−3 Cl + Br− (2.50)

R−+H + H3N+ − R′ − COOH = R−+NH3 − R′ − COOH + H+ (2.51)

Obviously, the limiting value for both ion exchange and non-exchange sorption ofthe solutes by reactions (2.48)–(2.51) is the ion exchange capacity. Number of non-stoichiometric interactions can take place instead or in addition to the stoichiometricprocesses. The two most common are sorption of electrolytes from concentrated solutionand sorption of large organic molecules by surface of the material. The electrolytes canbe transferred from the bulk solution to the internal solution of the ion exchange materialas pairs of cations and anions compensating the charge of each other.38 The transferred

38Such transfer is counteracted by ion exchangers and thus cannot be regarded as a simple distribution of thesubstance between two phases (see Section 7.7).


amount depends only on the concentration of the external solution and operating condi-tions. It is not limited by the amount of some groups or units contained in the material.As a result, the term capacity is not applicable to the process, however, the concepts ofutilised capacity and equilibrium capacity are used very often even omitting the words“utilised” and “equilibrium” (that is obviously a misleading practice). The surface sorp-tion is limited by the square of the pore surface. Thus the limiting value can be obtained.For example, if the sorption can be described by a Langmuir isotherm, the saturationof the isotherm corresponds to the sorption capacity. Like other adsorbents with activesurface, the surface sorption capacity of ion exchange materials is, among other factors,dependent on the size and morphology of the sorbed molecules. If the same substanceis sorbed by both ion exchange and non-exchange mechanisms, the total amount can bedefined as the over-all capacity [5].

Many sources operate with terms static ion exchange capacity or dynamic ion exchangecapacity. The words static and dynamic reflect practical methods used for the capacitydetermination, i.e. refer to the measurement method. As a matter of fact, static ionexchange capacity is a particular case of equilibrium capacity. The value is determinedat equilibrium with a significant extent of solution, i.e. when most of the available ionexchange groups are associated to (or can be assigned to) the target ion. While also beingessentially an equilibrium capacity, the dynamic capacity usually has a higher value andcan approach the total capacity because of the more efficient loading procedure.

Practical calculations often request a simplification of considered models. Concept ofapparent capacity or effective capacity can be useful in some cases. The approach takesinto consideration only counterions which can be exchanged in a certain process. Forexample, weak-acid or weak-base groups may be incompletely ionised and thus partlyinoperative. If the number of unionised groups remain constant during the process, thesegroups can be excluded from the capacity calculation. As a result, the apparent capacity,expressed in terms of exchangeable counterions, is strongly dependent on the experimen-tal conditions. Reduction of pH reduces ionisation degree of weak-acid groups. In case ofweak-base groups, the ionisation is suppressed by the pH increase. As a result, sorbentsbearing weak-acid groups can take up less, let us say, Na+ ions from acid than fromalkaline solutions. Correspondingly, weak-base exchangers can take up less Cl− ionsfrom alkaline than from acidic media. The apparent capacity of weak cation exchangersthus falls off when the pH drops below the ionisation constant (pK) of the groups. Thesame is true for anion exchangers when the pH rises above the pK of the groups [5].39

Another example could be an exchange of alkali metals on a chelating polymer partiallyloaded with a strongly complexing metal ion. For example, let us consider a poly-mer bearing picolinic groups with a total capacity of 4 meq/g of dry material. Let usassume that the material is partially loaded with copper: 1 meq of Cu2+ per gram of drymaterial. The exchanger sorbs copper ions almost irreversibly [157] due to the formationof extremely strong complexes [182]. No one alkali ion can compete with Cu2+ forthese groups. Hence, an exchange between K+ and Na+ can involve only functionality

39The pH inside ion exchangers differs from the pH of the external solution. Determination of pK values forfunctional groups is also a difficult task [143]. As a result, calculation of apparent capacity on the basis ofexperimental conditions could be difficult, if ever possible for many cases.


unoccupied with copper and, thus, the remaining 3 meq/g can be taken as the appar-ent capacity. Of course, its value cannot characterise either the material or its overallperformance but only its behaviour under certain conditions.

Ionisation degree of the sorbed substance can also affect the value of apparent capacity.For example, phosphoric acid can be exchanged depending on pH as PO3−

4 , HPO2−4 , or

H2PO−4 ions thus compensating three, two, or one charge of functional groups [5].

In many cases, the apparent capacity is limited by the size of the counterion[5,112,116,183–186]. For example, zeolites and highly cross-linked ion exchange resinsexclude counterions which are too large to fit into micropores of the material. Thisphenomenon is called sieve action [5].40 As a result of this phenomenon, the limitingcapacity for large organic ions depends on: cross-linking degree, size of the sorbed ions,etc. [116].

The complex dependence of the apparent capacity on the experimental conditions makesthis quantity rather unsuitable for characterisation of ion exchange materials and ismeaningful only when accompanied by a detailed description of the experiment (pHalone is insufficient). In some cases of weak exchangers, a better way of characterisingis to provide the total capacity and protonation constants for all involved groups [5].

Different capacity characteristics are interrelated; however, most of them are not constantbut depend on the operating conditions. As a result, recalculations between differentcapacity values for the same material could be difficult because of the need for additionalinformation about the measuring procedure. Any derivation of such relationships beginsfrom mathematical expression of the capacity definition. For example, the total capacity(Qtot) is expressed as

Qtot = nmd

meq/g (2.52)

where n is the number of equivalents of the available functional groups in the sample;md is the dry weight of the sample. The concentration of functional groups in the internalsolution (CR, N) is expressed as

CR = n

VW

(2.53)

where VW is the volume of the internal solution. Volume capacity of the exchanger bedin an ion exchange reactor is expressed as

Qbed = nVbed

meq/ml (2.54)

All other capacity characteristics can be expressed in a similar way. The interrelation canbe found considering the denominators in Eqs (2.52)–(2.54). For example, water content

40The sieve action will be discussed in detail in Section 4.1.


is swollen or air-dry material is usually expressed as percent of total weight. Hence, theamount of water in the sample can be calculated as

VW = ρW

W

100% − Wmd (2.55)

where ρW is the density of the internal solution inside the exchanger;41 W is the watercontent. Volume of the swollen sample (VR) relates to its dry weight as

VR = ρ100%

100% − Wmd (2.56)

where ρ is the density of the swollen material. VR is also connected with volume ofthe bed:

VR = Vbed100% − ε

100%(2.57)

where ε is the fractional void volume of the packing, i.e. fraction of the bed volumeunoccupied with the ion exchanger. Combining Eqs (2.52)–(2.57) could provide interre-lation between capacity characteristics Qtot , CR, and Qbed . However, one has to be surethat the meaning of n is the same in Eqs (2.52)–(2.54). Otherwise additional coefficientsor relationships must be introduced to indicate fraction of functional groups consideredin each case.

Derivations similar to (2.52)–(2.57) can be used to interrelate other capacity character-istics. The main difficulty is to secure that the measured values (like n in the exampleabove) are the same, i.e. the measuring conditions have affected the measuring of two(or more) capacity characteristics in the same way.

To conclude the section, the following warning must be expressed. The term ion exchangecapacity does not say much if there is no description of its meaning. In worst case(when no description is provided), one can suggest that the total ion exchange capacityis discussed but this suggestion is hardly reliable. Supplying a capacity value withcorresponding units is often insufficient because such dimensions as, for example, mil-liequivalents per gram or milliequivalents per cubic centimetre can refer to the swollenor dry material or to the exchanger bed.

2.7 Physical Structure

As described in Section 2.3, most of the organic ion exchange resins consist of anessentially irregular macromolecular, three-dimensional network of hydrocarbon chains

41Concentration of groups or ions in the internal solution is usually calculated with the assumption that thedensity of internal solution does not differ from the density of bulk solution of the same composition. Thisassumption is far from being precise because the internal liquid is a subject for many influences (elastic responseof the matrix, hydrophobic interactions with hydrocarbon chains, etc.).


bearing functional groups. The matrix and functional groups define the chemical proper-ties of the polymer. However, besides the chemical structure, physical configuration ofthe material is also highly important for the performance because it defines surface area,mechanical stability, resistance to a liquid flow, etc. Even thermodynamic and kineticcharacteristics of ion exchange are dependent on the macrostructure of the material (seeChapters 7–10).

Ion exchange polymers differ due to differently distributed density of the gel. Conven-tional, so called gel resins (a in Fig. 2.15) have relatively homogenous polymer densityacross the bead.42 The gels have no well-defined pore structure; however, the molecular-and nano-scale open areas between the hydrocarbon chains are conventionally designatedas pores without considering the true geometry.43 When the gels are swollen, these porescontain the solvent; the pores collapse at drying and are thus virtually non-existent indry state. The isoporous resins (b in Fig. 2.15) have specially designed homogenousdistribution of polymeric chains across the bead and thus more regular structure of themicropores.44

The most widely used materials nowadays are macroporous resins (c in Fig. 2.15).45

They have heterogeneous (in nano- and macro-scale) structure consisting of two phases:(i) gel regions containing dense polymer chains and a minor amount of the solvent,(ii) macroscopic permanent pores containing solution similar to the surrounding medium[189,190]. Such materials are obtained by synthesis in the presence of an inert diluent thatdissolves the monomers but precipitates the grooving copolymer. In the course of poly-merisation, the initially homogeneous mixture separates into two phases; one of which

(a,b) (c)

(a) (b) (c)

Fig. 2.15. Morphology (upper row) and distribution of polymeric chains (bottom row) in ionexchangers: (a) gel resins; (b) isoporous resins; (c) macroporous resins.

42Please note two meanings of the words conventional ion exchange resins. Traditionally, they refer to gel-typeion exchangers. However, the same words can also be used when well-known marketed materials are discussedwithout specifying their physical structure.43Only few inorganic exchangers have pores as channels of uniform cross section [187].44The difference between conventional gels and isoporous structures was also discussed in Section 2.3.45History of the origin and development of macroporous ion exchange resins can be found in the publicationof Abrams and Millar [188].


is the cross-linked polymer and the other is pure solvent (which defines the pore struc-ture) [140,191]. Macroporous resins have permanent pores with diameters of these poresin the range of 20–200 nm that is much larger in comparison with the distance betweenadjacent hydrocarbon chains of gel-type materials (0.5–20 nm) [187,190]. An importantfeature of macroporous materials is the relative constancy of their shape. Macroporesdo not collapse when these materials lose water, that is in contrary to pores of the gelswhich appear only in the swollen state (even their size is defined by the swelling degree).Macroporous resins are highly advantageous if the performance of the process is limitedby a slow diffusion of exchanged ions in the gel phase of the material. The diffusion ismuch faster when it takes place in the liquid phase filling the macropores (the solutionin macropores is the same or, at least, similar to the solution surrounding the exchanger).Besides the facilitation of the diffusion rates, the open-pore structure allows diffusion oflarge molecules. The exchange takes place on the surface of macropores or in close prox-imity to the surface, thus the molecules do not enter the dense gel regions. In additionto the faster process rates, macroporous ion exchange polymers exhibit better chemicalstability, especially resistance to oxidation [192].

Assembling ion recognition sites of imprinted polymers on the surface of the poreshas been successfully performed. Such polymers imprinted with metal ions have beenprepared by emulsion polymerisation under conditions allowing the formation of thedesired chelate complexes at the aqueous–organic interface of the emulsion. This posi-tioning of the functional sites has been preserved during the polymerisation process[193]. Imprinted polymers exhibit greatest affinity to desired ions and molecules. Thesematerials are discussed in detail in Section 3.2.

While surfing through literature on ion exchangers, one can find the term macroreticularresins. The term was introduced in an attempt to denote different types of porosity [188],however, it became a synonym for macroporous resins. Some authors use this term toemphasise the structure of beads consisting of well-defined micro-grains that is illustratedby Fig. 2.16.

Most of the ion exchange resins are marketed as beads or grains of spherical or undefinedshape. Non-spherical beads are often called blocks. The resins are also manufactured asmembranes, fibre-shaped materials, textiles, and other less conventional forms.

The granular ion exchange resins usually have a relatively wide distribution of the beadsize. However, many ion exchange processes and technologies impose certain limitationfor the bead size and often require use of monodisperse materials. As a result, manyapplications begin from a size separation of commercial ion exchangers that is mostoften performed with sieving. To provide customers with more convenience, most ofthe producers offer materials with narrow ranges of the size distribution of the beads.Special synthetic procedures allow syntheses of monodisperse materials. Figure 2.17shows morphology of the monodisperse material MonoPlus S 100 produced by Bayer.Figure 2.18 shows the size distribution diagram for this material in comparison with thecorresponding hetero-disperse polymer [195]. However, synthesis of mono-sized beadsis not a simple task and has been accomplished only in few cases. Thus most of themarketed materials with narrow bead-size distribution are simple sieved (or treated withsome other size-discriminating technique) prior to the packing.


microspheres

particleof resin

Fig. 2.16. Illustration of the term macroreticular resin. Micrograph of macroporous resinAmberlyst-15 (sulphonated polystyrene with 20% of divinylbenzene as the cross-linking agent)and corresponding schematic representation. Each bead consists of small, relatively uniform par-ticles separated by large pores. In fact, these particles are simply regions occupied with the highdensity polymer. The micrograph is reprinted from M. A. Harmer and Q. Sun, in: Applied CatalysisA - General [194] 2001, Elsevier ©.

Fig. 2.17. Monodisperse ion exchangers MonoPlus S 100 (Bayer). Reprinted from A. Scheffler,in: Ion Exchange at the Millennium [195] 2000, with kind permission of the Society of ChemicalIndustry ©.


0.0 0.2 0.4 0.6 0.8Bead size, mm

0

20

40

60

80

100

Sizedistribution

1.0 1.2

Fig. 2.18. Bead size distribution for monodisperse material MonoPlus S 100 (black curve) incomparison with corresponding hetero-disperse polymer (grey curve) [195].

The density of most of the polymeric ion exchangers is above 1 g/cm3. The densityof swollen cation exchangers is usually higher than the density of anion exchangers.Approximate values are: 1.10–1.40 g/cm3 for cation exchangers and 1.01–1.10 g/cm3

for anion exchangers [187]. This difference in density is often exploited for the separationof these two types of sorbents.46 Density of some weak anion exchange resins could bebelow the density of surrounding solution that makes certain inconvenience when thesematerials are used in packed bed reactors.

Specific applications require specific properties of ion exchange materials. For example,the expanded bed technique47 often requires higher density of the exchanger beads than

46Separation of a cation and anion exchange beads mixture is a common and regular task in so called mixedbed deionisation. The mixed bed technique will be discussed in Section 12.6.47The expanded bed technique will be discussed in Section 14.2.


that can be provided by the polymer itself. The solution is in the use of compositefunctional polymers incorporating heavy particles of inorganic nature. For example, theparticles of TiO2 have been integrated in cellulose-based materials supposed for anionexchange bioseparations in expanded bed columns [196]. Assemblies consisting of adense core, protecting polymeric layer, and the external layer of functional polymercould be even more advantageous due to the absence of functional groups located deepinside the bead. However, information about successful preparation of such materials islimited.

Many functional polymers are marketed nowadays in the form of fibres or textiles(Fig. 2.19). Inorganic fibrous ion exchangers consisting of monofilaments of uniformsize are also known (Fig. 2.20). The materials of fibrous structure have provided anumber of advantages:

• Short diffusion passes are predetermined, providing sorption rates that can be up tohundred times faster than that of the conventional granular resins (with a particlediameter usually between 0.25 and 1 mm) [160].

• Fibrous ion exchangers have higher osmotic stability that allows them to be mechan-ically more durable in conditions of multiple wetting and drying (for example, atcyclic sorption/regeneration processes in treatments of gaseous media) [160].

• They allow designing of packet reactors with pressure drops much lower than inreactors using granular materials [199].

Fig. 2.19. The SEM image of fibrous ion exchange polymer bearing amidoxime groups (left) andthe same material loaded with Au3+ (right). Reprinted from W. P. Lin, Y. Lu, and H. M. Zeng,in: Journal of Applied Polymer Science [197] 1993, with kind permission of John Wiley &Sons Inc. ©.


Fig. 2.20. Electron micrograph of an inorganic fibrous ion exchanger polyacrylonitrilethorium (IV) phosphate consisting of monofilaments of uniform size ranging between 5 and 50 µm.Reprinted from K. G. Varshney, N. Tayal, A. A. Khan, and R. Niwas, in: Colloids and SurfacesA - Physicochemical and Engineering Aspects [198] 2001, Elsevier ©.

20mm

Fig. 2.21. Morphology of fibrous cation exchanger (sulphonated polystyrene) obtained with elec-trospray deposition. Reprinted from H. Matsumoto, Y. Wakamatsu, M. Minagawa, and A. Tanioka,in: Journal of Colloid and Interface Science [200] 2006, Elsevier ©.

• Fibrous ion exchangers can be used in the form of various textile goods such ascloth, conveyer belts, non-woven materials, staples, nets, etc. This opens manyunusual possibilities for new technological designs [160].

Novel techniques allow preparation of ion exchangers with unusual morphology. Forexample, organic fibrous cation and anion exchangers shown in Fig. 2.21 have beenobtained with electrospray deposition [200]. Another composite material is shown inFig. 2.22 [201]. It has been prepared by coating of a glass fibre substrate with a styrene–divinylbenzene oligomer and then by curing and sulphonation. Figure 2.23 presents one


Fig. 2.22. Composite ion exchange material consisting of glass fibres covered with sulphonatedstyrene–divinylbenzene polymer. Reprinted from L. Dominguez, K. R. Benak, and J. Economy,in: Polymers for Advanced Technologies [201] 2001, with kind permission of John Wiley & SonsLimited ©.

Bonding by melting of binder

Ceramic fibre

Fig. 2.23. Composite zeolite sheets based on a mixture of cotton linters pulp, ceramic fibre, andzeolite. Alumina sol was used as the binder. The sheets were ignited at 700◦C. Reprinted fromH. Ichiura, N. Okamura, T. Kitaoka, and H. Tanaka, in: Journal of Materials Science [202] 2001,with kind permission of authors and Springer Science and Business Media ©.


more example of ion exchange composite based on a mixture of cotton linters pulp,ceramic fibre, and zeolite [202].

Hypercrosslinked polymeric networks [140,144], particularly Macronet Hypersol MN-series (Purolite Int. Ltd.) are characterised by a very high internal surface area (1000 m2/gor higher) [72,140,146] while the surface of conventional macroporous ion exchangeresins is an order of magnitude less. Scanning electron microscopy examination of suchpolymers reveals a uniform globular structure (the globule diameter 120–160 nm) for allthe Macronets. The large inter-globular pores enable rapid diffusion of sorbates. Textureof the surface indicates possible presence of macropores with size around 20 nm [146].

Another class of ion exchange materials is membranes. Their chemical structure is thesame as the structure of ion exchange resins: cross-linked polymer chains bearing eitherfixed positive or fixed negative charges. Membranes are manufactured as continuoussheets, tubes, disks, ribbons, etc. Manufacturing methods define two types of membranes:homogeneous and heterogeneous membranes. Homogeneous membranes are more com-mon nowadays. They are ion exchange gels with ionic groups evenly distributed in themembrane bulk, i.e. they consist of conventional gel-type ion exchange resins. Heteroge-neous membranes contain particles of an ion exchanger incorporated in an inert support(polystyrene, polyethylene, wax, etc., see Fig. 2.24). They possess a high mechanicalstability but their performance in different practical applications is not as good as those

+

−−

−

+

+Inert binder

Cation exchanger

Fig. 2.24. Heterogeneous cation exchange membrane. Particles of the exchanger are embeddedin an inert binder. The direct contact between beads of the ion exchanger is essential to allowfree migration of counterions in, out, and through the membrane. Co-ions are repelled by thefixed charges (as described later in Section 7.7) and cannot pass through the membrane. Inertbinder does not allow passage for any ions. As a result, only positively charged ions can migratethrough cation exchange membranes, while anion exchange membranes allow the passage of onlynegatively charged ions [71]. The pass-ways are torturous and defined by packing of the exchangerparticles.


Table 2.4. Main properties of ion exchange resins.

Chemical Physical

• Type of the matrix • Physical structure and morphology• Cross-linking degree • Particle size• Type of functional groups • Pore size and morphology• Ion exchange capacity • Surface area• Ionic form • Partial volume in swollen state

of the homogeneous membranes. Heterogeneous membranes are not widely used nowa-days, however, they are “easy to prepare” and thus are convenient laboratory test objects.More about properties of ion exchange membranes and exploitation of these propertiesfor separation needs is discussed in Chapter 17.

2.8 Properties of Ion Exchange Resins – Summary

Properties of ion exchange resins have to be summarised once more to conclude thischapter. The list of main properties is presented in Table 2.4. Of course, each materialhas many more characteristics. The list contains characteristics that are most commonfor all ion exchange materials. It allows to identify the material while its performance indifferent processes could be described with the use of corresponding secondary values.For example, chemical affinity towards different types of ions is characterised by differ-ent selectivity characteristics (see Section 8.4). However, the affinity is a product of acombination between the type of functional groups, type of matrix, cross-linking degree,ion exchange capacity, swelling, and even composition of the surrounding media.

Documents

Ion Exchange Materials || Ion Exchangers, their Structure and Major Properties